Our research addresses two aims: 1) to characterize the psychobiological correlation of response to CBT and 2) to clarify the state dependence of selected neurobiological disturbances acorss the course of the illness. We have found that clinical severity, an abnormal EEG sleep profile and increased HPA activity are associate with poorer response to DBC which suggests that some unipolar depressed patients may require somatic treatment in order to correct/control dysregulated CNS processes. We propose to extend this work in a controlled study of 200 depressed individuals.